Syntesis of carbon nanostructures near room temperature using ...
1
Syntesis of carbon nanostructures near room temperature using PECVD assisted by microwave Flavio Carvalho1, Alfredo A. Vaz2, Mário A. Bica de Moraes2, Stanislav Moshkalev2, Rogério V. Gelamo1* 1
Universidade Federal do Triângulo Mineiro, Uberaba, MG., Brazil, 38064-200 2
Universidade Estadual de Campinas, Campinas, SP., Brazil, 13083-870 *
[email protected]
ABSTRACT Carbon nanostructures (nanofibers, nanosponges and nanospheres) were synthesized onto Si (001) substrates using plasma enhanced chemical vapor deposition (PECVD) from C2H2-Ar mixtures. The plasma was activated by a microwave generator and the nanostructures were grown at low substrate temperatures (about 120 ºC) on top of previously deposited catalytic Ni and Cu films of 3 nm thick. No films treatment was made prior to the deposition of the carbon nanostructures. Nanoholes, nanospheres and nanofibers were obtained, depending on acetylene partial pressure used during depositions. Atomic force microscopy (AFM) and scanning electron spectroscopy (SEM) were employed for the morphological characterization of the catalytic films, in order to investigate carbon nanostructures growth mechanisms. Raman spectroscopy was used to investigate carbon hybridization states. Nanofibers of 300 to 500 nm long were observed for some plasma conditions (pressure and microwave power), while a mixture of sp2 and sp3 hybridizations were revealed by the Raman spectra. The results shown here indicate a promising simple and low cost technique for the production of conductive carbon nanofibers connected to Si wafers.
2
Keywords: nanofiber, PECVD, microwave plasma, acetylene, AFM, SEM, Raman spectroscopy, carbon hybridization states
1. Introduction Since the discovery of a third allotropic form of carbon [1], and the developments in the synthesis of carbon nanotubes [2], investigations in the synthesis of nanostructured materials have dramatically increased. As a consequence, the use of carbon on new materials [3] and composites [4] has grown steadily. Such carbon structures can be synthesized by the technique of plasma enhanced chemical vapor deposition (PECVD) and can be exemplified by diamond-like carbon [5], carbon nanofibers [6], carbon nanotubes [7], amongst other forms of carbon [8]. PECVD is indeed of great interest, as depositions can be carried out at relatively low substrate temperatures, thus avoiding the sometimes inconvenient temperature effects in the deposited material. Furthermore, owing to the usually high deposition rate of PECVD [9, 10, 11, 12, 13], the structures can be formed at short times, which is of interest in the industrial point of view. In the PECVD technique, the carbon nanostructures strongly depend on the deposition parameters. Is possible to observe subtle variations between the parameters for obtaining carbon nanofibers in works that relate the obtaining of these nanostructures using PECVD near room temperatures. The differences being focused on the type of diluent gas and carbon source and control of font type plasma generated. Chen and colleagues [9] using dilute hydrogen to methane as key to their synthesis of nanofibers, Choi et all. [10], using a pre-treatment with hydrogen plasma at different intensities before admission of the carbon source in the synthesis chamber. Wang and Moore [11] provide the use of hydrogen for the synthesis of nanofibers, using it only for nanotubes, focusing on the control of synthesis of nanofibers on the intensity of the argon
3
plasma, with ignition at 30 W. Hoffmann et all. [12] using ammonia plasma for obtaining the nanofibers together with the carbon source, resulting in a lower deposition temperature lower of the order of 120º C . Regarding carbon nanofibres, they are formed from a catalytic process involving ultrafine particles (few nm-diameter) of Ni, Cu or another catalyst. Growth occurs through an epitaxial process starting with the adsorption of carbon on the catalyst surface, originating from hydrocarbon decomposition, either in a glow discharge or by a thermal process [19]. According to the literature, Ni is widely used as a catalyst for nanotube and nanofiber growth. Their structure can be controlled according to the catalyst pretreatment. With the use of radio frequency plasmas, nanofibers and nanotubes [11, 13] were obtained in various different conditions (different precursors, gas pressures, substrate temperatures). Microwave plasmas have also been used for growing nanofibers [20], nanotubes [10] and nanosheets [22], however to get carbon nanostructures, substrate temperatures using microwaves as energy source (450 to 600 o
C.) are usually higher than those in depositions carried out in dc or rf plasma assisted
discharges. The deposition of carbon nanostructures, nanotubes and nanofibres mainly using plasma processes, have been shown promising especially with respect to lower deposition temperatures compared to the techniques of CVD and arc discharge. The type and shape of the nanoparticle catalyst also deserves more research to reach carbon structures with fewer defects to facilitate their integration in Si-based devices .In this work we report an investigation on the growth of nanoholes, nanospheres and nanofibers, using a microwave PECVD system of in-house design. All nanostructures were formed onto Si (001) substrates coated with Ni or Cu films. Depositions were performed close to room temperatures (120 ºC at most); no pretreatment of the catalytic film was made. The morphology of the structures was characterized by SEM and AFM as a
4
function of the acetylene partial pressure and metal catalysts, while chemical bonding was investigated by Raman spectroscopy.
2. Experimental In this work carbon nanostructures were synthesized by PECVD on Si coated with catalysts thin films previously deposited. The substrates [Si (001), 1 x 1 cm2, were cleaned at the RCA Cleaning Center of the Center of Semiconductor Components (CCS) of Unicamp]. Using a dc magnetron sputtering system, working at a base pressure of 1 x 10-7 Torr, Ni and Cu catalytic films were deposited upon Si waffer, whose thickness (~3 nm) was measured with a quartz microbalance. Figure 1 is a schematic representation of the PECVD system. Vacuum is produced with a rotary vane pump of 10 m3/h pumping speed. For the generation of the plasma, a microwave source (4.5 GHz, 500 Watt) was employed. Acetylene (C2H2, 99,99% pure) was used as the precursor, mixed with argon (99,99% purity). The pressure ratio of Ar to C2H2 was kept constant and the residual pressure before gases admittance was 60 mTorr. During depositions, pressure and substrate temperature were measured with a Pirani gauge and with a type-K thermocouple. The morphology of carbon nanostructures surfaces, were an by scanning electron microscopy, Dual Beam FIB/SEM (Focused Ion Beam/Scanning Electron Microscopy) Model Nanolab Nova 200 (FEI Co.) (FIB / FEG) model New Nanolab 200 (FEI Co.) installed at Center for Semiconductor Research/UNICAMP. All images were made using an electron beam at 5 kV and 0.4 nA. The secondary electron mode was used to obtain surface images. To estimate of the average diameters and height of the nanostructures and their standard deviations we used the SEM images in the monitor screen, a number of particles were chosen and their sizes individually measured and stored to posterior calculations.
5
To analyze the morphology of the silicon substrate and the Cu and Ni films, AFM images in dynamic mode were obtained, using the Shimadzu microscope model SPM9700, installed in the UFTM. Obtaining of the average roughness of the Si and Ni and Cu films were made using the 9700 Scanning Probe Microscopy software provided by Shimadzu Corporation. Chemical bonding in the carbon nanostructures was investigated using a confocal Raman Spectroscope, model Ntegra Spectra from NT MDT.Co of CCS-Unicamp. The 473.8 nm line of a semiconductor laser was used for excitation.
3. Results and Discussion Various samples, deposited in different conditions and on either Ni or Cu catalysts (~3nm thickness) are investigated in this work. As shown in Table I, samples are labeled by a number (the C2H2 partial pressure in mTorr), followed by the catalyst chemical symbol. The catalyst and its thickness, the deposition parameters of the PECVD system and the carbon structures observed are also given in the table. Figure 2 shows AFM pictures of the surfaces of the bare Si substrate, and the 3 nm Cu and Ni films. An average surface roughness of 0.178 nm is measured for the Si surface, typical of Si wafers. For the Cu and the Ni films, the average roughness is 0.358 and 0.227 nm, respectively. As revealed by the pictures, the films consist of metal particles which cover the substrate, i. e., in general there are no isolated metal islands on top of the substrate. Figures 2b and 2c, are representative of all Cu and Ni films of this work, since the catalysts were deposited by sputtering on a waffer of Si and subsequently cut into pieces of 1x1cm. Different nanostructures were obtained depending on the partial pressure of C2H2 used in the deposition. The lowest one (34 mTorr), is responsible for the formation, on top of the Ni film (Figure 3A), whose surface is shown consisting of nearly spherical fine particles of relatively
6
uniform diameter (41.8 ± 4.9 nm). On the other hand, nanoholes of a rather large size distribution are seen in the film grown over the Cu catalyst (Figure 3B), presenting a average diameter of 308.3 ± 56.4 nm. Nanoholes are also observed in films grown over both catalysts at the pressures of 64 mTorr (samples 64Ni and 64Cu) for both catalysts observed in Figure 4, and average diameter of 275.6 ± 77.2 and 216.6 ± 40.1 nm, respectively. The early stages of formation of nanotubes with nanospheres in the tip of the tubes are shown in Figure 5 for samples 70Ni and 70Cu (Figures 5A and 5B, respectively). The average diameter of nanospheres is 56.3 ± 12.1 nm to 70Ni and 46.4 ± 5.6 nm to 70Cu. With 100 mTorr of acetylene, a nano spongy structure is obtained with some nucleation sites tubes arranged along the surface of the films, as showed in Figure 6. At the pressure of 79 mTorr of acetylene, a huge number of carbon nanotubes are formed as shown in Figure 7. They probable grow from the bottom of the nanoholes and more than one nanotube grows from each nanohole. From the Figure 7 we can readily suppose that as each nanotube grows, it laterally touches other nanotubes rom the same hole, forming nanotube bundles which in turn touch other adjacent bundles, resulting, in some regions of the film, in a very dense sponge-like structure. The average height of nanofibers is 343.0 ± 32 nm to 79Ni sample and 381.4 ± 16.5 nm to 79Cu sample. Average diameter of nanofibers estimated in the half of the tubes is 46.7 ± 13.6 nm to 79Ni and 27.9 ± 7.9 nm to 79Cu sample, respectively. It is interesting to compare our results with those of other investigators in which PECVD was also used to obtain carbon nanostructures and external power sources were used for substrate heating. Hoffman at al. [25] have grown carbon nanofibers over a Si substrate decorated with micrometric patches of 8.0 nm thick Ni films by a lithographic process. They used an external power source to heat the substrate at 120 °C. Wang et al. [11] Obtained carbon nanofibers at 140 °C over 8.0 nm Ni films, also using an external power source for heating. In
7
their work the nanofibers have length about 600.0 nm, but presented rough edges and pores. Our method of nanostructure synthesis, therefore, has the advantage that no thermal treatment of the catalysts before deposition is required, and there is no need of an external power source for substrate heating. These simplifications are of interest in the technological point of view. According to the SEM images, particularly in Figure 7, due to no pre-treatment of the film catalyst, it is possible to estimate that the nanofibers showed in our work have a “base growth” mechanism [11, 19, 23]. Wang and Moore [11] showed that due to the deposition process of Fe and Ni catalyst nanoparticle by sputtering, the base growth process is favored as the catalyst nanoparticle has greater adhesion to glass or Si substrate due to the formation of bonds between them. As the growth process takes place at temperatures of about 100oC , the nanoparticle does not melt and nanofibre growth occurs by diffusion and dissociation of carbon in the metal, this occurring at lower energy because the carbon is largely dissociated in plasma gas phase. As can be seen in SEM images that the nanofibers have a conical shape of approximately 500.0 nm in height with a base diameter smaller than top, suggesting that the break of growth occurs on the substrate (Ni and Cu nanoparticle). In this case, the mechanism of nanofibre growth is initiated by the adsorption of the monomer on the surface of the film catalyst, which is a continuous film, as can be seen in the AFM images of the catalysts films (Figure 2), which hinders the migration of the particle to the top of the nanofiber, remaining fixed on the base, in the substrate. With the nanofibers growth, the diffusion of carbon through the metal surface decreases along with the minimum energy for catalysis, prevented by the barrier created by nanofibers, this barrier increases with the height of the conical shape nanofibers generated culminating in the disruption of growth observed in the base.
8
Another factor that may contribute to the stop of nanofiber growth is the low diffusion of carbon or carbon-containing species generated in the plasma discharge on the surface of the metal catalyst due of certain quantity of amorphous carbon structure formed between nanofibers during the plasma discharge. As can be observed in Figure 7, the fibers and the amorphous carbon film grows and forms a dense structure that can act as a barrier to carbon permeation, stopping fiber growth and contributing to the smaller diameter of the fibers at their basis. The carbon nanostructures were further investigated by Raman spectroscopy. Figure 8 shows a Raman spectrum of sample 79Ni which is representative of the spectra of 79Cu sample. Pronounced peaks are seen, readily identified as the D (disorder) and G (graphitic) bands, at 1360 and 1610 cm-1, respectively. The high intensity of the G band is taken as an evidence of the large sp2 carbon density in the fibers, while the overlapping of the G and D bands is interpreted as a mixture of sp2 and sp3 bonding states. The carbon nanofibers showed in image (Figure 7A), which are basically defective nanotubes have sp2 carbons in its crystal lattice demonstrated in 1587 cm-1 (G band) with an higher intensity than the shoulder formed by the D band at 1370 cm-1 indicating a combination of sp3 carbons in the nanostructure. But in this case, the linewidths is large to compare with crystalline 2D graphite due presence of residual amorphous carbon. In fact our nanofibers exhibited a mixture between crystalline and amorphous carbon in the structure.
4. Conclusions In this work we have shown that the use of microwave-assisted PECVD can be a successful technique for the synthesis of carbon nanotubes. The various acetylene partial pressures we have used facilitate the observation of the morphology of the carbon nanoholes and nanotubes in their various stages of formation.
9
The very well-defined images of the carbon nanotubes and of their coalescence provide strong evidence for the base growth mechanism proposed in other investigations. From the Raman spectrum, the sp2-nature of the carbon nanotubes and the existence of mixed sp2-sp3 bond states was clearly evidenced. The PECVD system we have used is rather simple and inexpensive. Furthermore, the procedure adopted does not require any elaborate previous substrate treatment such as thermal annealing or holographic processes. Acknowledge The authors thank to CNPq and FAPEMIG for the financial support to this work.
References
1.
Kroto, H W., Heath J. R., O’brien, J. C., Curt, R. F. & Smalley, R. E. C60:Buckminsterfullerene. Nature 318, 162–163 (1985).
2.
Iijima, S. Helical microtubules of graphitic carbon. Nature 354, 56–58 (1991).
3.
Herbst, M. H. Tecnologia dos Nanotubos de carbono: tendências e perspectivas de uma área multidisciplinar. Química nova 27, 986–992 (2004).
4.
Stankovich, S. et al. Graphene-based composite materials. Nature 442, 282–286 (2006).
5.
Grill, A. Diamond-like carbon: state of the art. Diamond and Related Materials 8, 428– 434 (1999).
10
6.
Melechko, a. V. et al. Control of carbon nanostructure: From nanofiber toward nanotube and back. Journal of Applied Physics 102, 074314–2–074314–7 (2007).
7.
Ren, Z. F. Synthesis of Large Arrays of Well-Aligned Carbon Nanotubes on Glass. Science 282, 1105–1107 (1998).
8.
Pham-Huu, C. et al. About the octopus-like growth mechanism of carbon nanofibers over graphite supported nickel catalyst. Journal of Catalysis 240, 194–202 (2006).
9.
Chen, Y., Guo, L. ., Johnson, D. . & Prince, R. . Plasma-induced low-temperature growth of graphitic nanofibers on nickel substrates. Journal of Crystal Growth 193, 342–346 (1998).
10.
Choi, W. S., Choi, S.-H., Hong, B. & Lee, J.-H. Growth of carbon nanotubes on glass substrate by MPECVD. Materials Science and Engineering: C 26, 1215–1218 (2006).
11.
Wang, H. & Moore, J. J. Low temperature growth mechanisms of vertically aligned carbon nanofibers and nanotubes by radio frequency-plasma enhanced chemical vapor deposition. Carbon 50, 1235–1242 (2012).
12.
Hofmann, S., Ducati, C., Robertson, J. & Kleinsorge, B. Low-temperature growth of carbon nanotubes by plasma-enhanced chemical vapor deposition. Applied Physics Letters 83, 135–137 (2003).
13.
Löffler, R. et al. Optimization of plasma-enhanced chemical vapor deposition parameters for the growth of individual vertical carbon nanotubes as field emitters. Carbon 49, 4197– 4203 (2011).
11
14.
Faria, A. & Gomes, N. Filmes de nanodiamantes para aplicações em sistemas eletroquímicos e tecnologia aeroespacial. Química nova 29, 129–136 (2006).
15.
Moshkalev, S. a. & Verissimo, C. Nucleation and growth of carbon nanotubes in catalytic chemical vapor deposition. Journal of Applied Physics 102, 044303 (2007).
16.
Guo, T. et al. Self-Assembly of Tubular Fullerenes. Journal of Physics and Chemistry 99, 10694–10697 (1995).
17.
Anderson, T. T. et al. A plasma arc reactor for fullerene research. Review of Scientific Instruments 65, 3820 (1994).
18.
Lee, B. J. & Kim, J. Recent Progress in the Synthesis of Porous Carbon Materials. advanced materials 18, 2073–2094 (2006).
19.
Dresselhauss, M. S., Dresselhauss, G. & Avouris, P. Carbon Nanotubes: Synthesis, Structure, Properties and Applications. (Springer: 2001).
20.
Hoshi, F. et al. Field emission and structure of aligned carbon nanofibers deposited by ECR-CVD plasma method. Diamond and Related Materials 10, 254–259 (2001).
21.
Kolahdouz, Z. et al. Substrate engineering for Ni-assisted growth of carbon nano-tubes. Materials Science and Engineering: B 177, 1542–1546 (2012).
22.
Wang, Z., Shoji, M. & Ogata, H. Carbon nanosheets by microwave plasma enhanced chemical vapor deposition in CH4–Ar system. Applied Surface Science 257, 9082–9085 (2011).
12
23.
Baker, R. T. K., Barber, M. A., Harris, P. S., Feates, F. S. & Waite, R. J. Nucleation and Growth of Carbon Deposits from the Nickel Catalyzed Decomposition of Acetylene. Journal of catalysis 62, 51–62 (1972).
24.
Boskovic, B. O., Stolojan, V., Khan, R. U. a, Haq, S. & Silva, S. R. P. Large-area synthesis of carbon nanofibres at room temperature. Nature materials 1, 165–8 (2002).
25.
Hofmann, S., Kleinsorge, B., Ducati, C. & Robertson, J. Controlled low-temperature growth of carbon nanofibres by plasma deposition. New Journal of physics 5, 153.1–153.3 (2003).
26.
Tuinstra, F. Raman Spectrum of Graphite. The Journal of Chemical Physics 53, 11261130 (1970).
27.
Dresselhaus, M. S., Dresselhaus, G., Saito, R. & Jorio, a. Raman spectroscopy of carbon nanotubes. Physics Reports 409, 47–99 (2005).
28.
Ferrari, a. & Robertson, J. Interpretation of Raman spectra of disordered and amorphous carbon. Physical Review B 61, 14095–14107 (2000).
29.
Ayoub, S. & Beaulieu, L. Y. The surface morphology of thin Au films deposited on Si(001) substrates by sputter deposition. Thin Solid Films 534, 54–61 (2013).
13
TABLE CAPTION Tabel I: Deposition parameters and catalyst metal for each sample. For all samples: deposition time of 20 minutes; pressure of the C2H2-Ar mixture constant at 210 mTorr; maximum substrate temperature: of 120 ºC.
FIGURE CAPTIONS Figure 1: Schematic drawing of PECVD system. Figure 2: AFM images of Si waffer and Cu and Ni catalyst films used . Figure 3: SEM images of structures deposited with 34 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 4: SEM images of structures deposited with 64 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 5: SEM images of structures deposited with 70 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 6: SEM images of structures deposited with 100 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 7: SEM images of structures deposited with 79 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 8: Raman spectra of nanofibers deposited with 79 mTorr of acetylene upon 3nm Ni.
14
TABLE 1 Sample
Catalyst
Catalyst
Argon
Thickness pressure
Acetylene
Structure
pressure
observed
(nm)
(mtorr)
(mtorr)
34Ni
Ni
3
110
34
Nanocones
34Cu
Cu
3
110
34
Nanosponge
64Ni
Ni
3
70
64
Nanoholes
64Cu
Cu
3
70
64
Nanoholes
70Ni
Ni
3
70
70
Nanospheres
70Cu
Cu
3
70
70
Nanospheres
79Ni
Ni
3
60
79
Nanofibers
79Cu
Cu
3
60
79
Nanofibers
84Ni
Ni
3
50
84
Nanoholes
84Cu
Cu
3
50
84
Nanoholes
15
FIGURE 1
MICROWAVE SOURCE
NEEDLE VALVES
QUARTZ CHAMBER
VACUUM PUMP
GAS SOURCE
16
FIGURE 2
17
FIGURE 3
A
B
18
FIGURE 4
A
B
19
FIGURE 5
A
B
20
FIGURE 6
A
B
21
FIGURE 7
A
B
22
FIGURE 8
G Intensity (arb. units)
D
1000
1200
1400
1600
1800 -1
Wavenumber (cm )
2000
Syntesis of carbon nanostructures near room temperature using PECVD assisted by microwave Flavio Carvalho1, Alfredo A. Vaz2, Mário A. Bica de Moraes2, Stanislav Moshkalev2, Rogério V. Gelamo1* 1
Universidade Federal do Triângulo Mineiro, Uberaba, MG., Brazil, 38064-200 2
Universidade Estadual de Campinas, Campinas, SP., Brazil, 13083-870 *
[email protected]
ABSTRACT Carbon nanostructures (nanofibers, nanosponges and nanospheres) were synthesized onto Si (001) substrates using plasma enhanced chemical vapor deposition (PECVD) from C2H2-Ar mixtures. The plasma was activated by a microwave generator and the nanostructures were grown at low substrate temperatures (about 120 ºC) on top of previously deposited catalytic Ni and Cu films of 3 nm thick. No films treatment was made prior to the deposition of the carbon nanostructures. Nanoholes, nanospheres and nanofibers were obtained, depending on acetylene partial pressure used during depositions. Atomic force microscopy (AFM) and scanning electron spectroscopy (SEM) were employed for the morphological characterization of the catalytic films, in order to investigate carbon nanostructures growth mechanisms. Raman spectroscopy was used to investigate carbon hybridization states. Nanofibers of 300 to 500 nm long were observed for some plasma conditions (pressure and microwave power), while a mixture of sp2 and sp3 hybridizations were revealed by the Raman spectra. The results shown here indicate a promising simple and low cost technique for the production of conductive carbon nanofibers connected to Si wafers.
2
Keywords: nanofiber, PECVD, microwave plasma, acetylene, AFM, SEM, Raman spectroscopy, carbon hybridization states
1. Introduction Since the discovery of a third allotropic form of carbon [1], and the developments in the synthesis of carbon nanotubes [2], investigations in the synthesis of nanostructured materials have dramatically increased. As a consequence, the use of carbon on new materials [3] and composites [4] has grown steadily. Such carbon structures can be synthesized by the technique of plasma enhanced chemical vapor deposition (PECVD) and can be exemplified by diamond-like carbon [5], carbon nanofibers [6], carbon nanotubes [7], amongst other forms of carbon [8]. PECVD is indeed of great interest, as depositions can be carried out at relatively low substrate temperatures, thus avoiding the sometimes inconvenient temperature effects in the deposited material. Furthermore, owing to the usually high deposition rate of PECVD [9, 10, 11, 12, 13], the structures can be formed at short times, which is of interest in the industrial point of view. In the PECVD technique, the carbon nanostructures strongly depend on the deposition parameters. Is possible to observe subtle variations between the parameters for obtaining carbon nanofibers in works that relate the obtaining of these nanostructures using PECVD near room temperatures. The differences being focused on the type of diluent gas and carbon source and control of font type plasma generated. Chen and colleagues [9] using dilute hydrogen to methane as key to their synthesis of nanofibers, Choi et all. [10], using a pre-treatment with hydrogen plasma at different intensities before admission of the carbon source in the synthesis chamber. Wang and Moore [11] provide the use of hydrogen for the synthesis of nanofibers, using it only for nanotubes, focusing on the control of synthesis of nanofibers on the intensity of the argon
3
plasma, with ignition at 30 W. Hoffmann et all. [12] using ammonia plasma for obtaining the nanofibers together with the carbon source, resulting in a lower deposition temperature lower of the order of 120º C . Regarding carbon nanofibres, they are formed from a catalytic process involving ultrafine particles (few nm-diameter) of Ni, Cu or another catalyst. Growth occurs through an epitaxial process starting with the adsorption of carbon on the catalyst surface, originating from hydrocarbon decomposition, either in a glow discharge or by a thermal process [19]. According to the literature, Ni is widely used as a catalyst for nanotube and nanofiber growth. Their structure can be controlled according to the catalyst pretreatment. With the use of radio frequency plasmas, nanofibers and nanotubes [11, 13] were obtained in various different conditions (different precursors, gas pressures, substrate temperatures). Microwave plasmas have also been used for growing nanofibers [20], nanotubes [10] and nanosheets [22], however to get carbon nanostructures, substrate temperatures using microwaves as energy source (450 to 600 o
C.) are usually higher than those in depositions carried out in dc or rf plasma assisted
discharges. The deposition of carbon nanostructures, nanotubes and nanofibres mainly using plasma processes, have been shown promising especially with respect to lower deposition temperatures compared to the techniques of CVD and arc discharge. The type and shape of the nanoparticle catalyst also deserves more research to reach carbon structures with fewer defects to facilitate their integration in Si-based devices .In this work we report an investigation on the growth of nanoholes, nanospheres and nanofibers, using a microwave PECVD system of in-house design. All nanostructures were formed onto Si (001) substrates coated with Ni or Cu films. Depositions were performed close to room temperatures (120 ºC at most); no pretreatment of the catalytic film was made. The morphology of the structures was characterized by SEM and AFM as a
4
function of the acetylene partial pressure and metal catalysts, while chemical bonding was investigated by Raman spectroscopy.
2. Experimental In this work carbon nanostructures were synthesized by PECVD on Si coated with catalysts thin films previously deposited. The substrates [Si (001), 1 x 1 cm2, were cleaned at the RCA Cleaning Center of the Center of Semiconductor Components (CCS) of Unicamp]. Using a dc magnetron sputtering system, working at a base pressure of 1 x 10-7 Torr, Ni and Cu catalytic films were deposited upon Si waffer, whose thickness (~3 nm) was measured with a quartz microbalance. Figure 1 is a schematic representation of the PECVD system. Vacuum is produced with a rotary vane pump of 10 m3/h pumping speed. For the generation of the plasma, a microwave source (4.5 GHz, 500 Watt) was employed. Acetylene (C2H2, 99,99% pure) was used as the precursor, mixed with argon (99,99% purity). The pressure ratio of Ar to C2H2 was kept constant and the residual pressure before gases admittance was 60 mTorr. During depositions, pressure and substrate temperature were measured with a Pirani gauge and with a type-K thermocouple. The morphology of carbon nanostructures surfaces, were an by scanning electron microscopy, Dual Beam FIB/SEM (Focused Ion Beam/Scanning Electron Microscopy) Model Nanolab Nova 200 (FEI Co.) (FIB / FEG) model New Nanolab 200 (FEI Co.) installed at Center for Semiconductor Research/UNICAMP. All images were made using an electron beam at 5 kV and 0.4 nA. The secondary electron mode was used to obtain surface images. To estimate of the average diameters and height of the nanostructures and their standard deviations we used the SEM images in the monitor screen, a number of particles were chosen and their sizes individually measured and stored to posterior calculations.
5
To analyze the morphology of the silicon substrate and the Cu and Ni films, AFM images in dynamic mode were obtained, using the Shimadzu microscope model SPM9700, installed in the UFTM. Obtaining of the average roughness of the Si and Ni and Cu films were made using the 9700 Scanning Probe Microscopy software provided by Shimadzu Corporation. Chemical bonding in the carbon nanostructures was investigated using a confocal Raman Spectroscope, model Ntegra Spectra from NT MDT.Co of CCS-Unicamp. The 473.8 nm line of a semiconductor laser was used for excitation.
3. Results and Discussion Various samples, deposited in different conditions and on either Ni or Cu catalysts (~3nm thickness) are investigated in this work. As shown in Table I, samples are labeled by a number (the C2H2 partial pressure in mTorr), followed by the catalyst chemical symbol. The catalyst and its thickness, the deposition parameters of the PECVD system and the carbon structures observed are also given in the table. Figure 2 shows AFM pictures of the surfaces of the bare Si substrate, and the 3 nm Cu and Ni films. An average surface roughness of 0.178 nm is measured for the Si surface, typical of Si wafers. For the Cu and the Ni films, the average roughness is 0.358 and 0.227 nm, respectively. As revealed by the pictures, the films consist of metal particles which cover the substrate, i. e., in general there are no isolated metal islands on top of the substrate. Figures 2b and 2c, are representative of all Cu and Ni films of this work, since the catalysts were deposited by sputtering on a waffer of Si and subsequently cut into pieces of 1x1cm. Different nanostructures were obtained depending on the partial pressure of C2H2 used in the deposition. The lowest one (34 mTorr), is responsible for the formation, on top of the Ni film (Figure 3A), whose surface is shown consisting of nearly spherical fine particles of relatively
6
uniform diameter (41.8 ± 4.9 nm). On the other hand, nanoholes of a rather large size distribution are seen in the film grown over the Cu catalyst (Figure 3B), presenting a average diameter of 308.3 ± 56.4 nm. Nanoholes are also observed in films grown over both catalysts at the pressures of 64 mTorr (samples 64Ni and 64Cu) for both catalysts observed in Figure 4, and average diameter of 275.6 ± 77.2 and 216.6 ± 40.1 nm, respectively. The early stages of formation of nanotubes with nanospheres in the tip of the tubes are shown in Figure 5 for samples 70Ni and 70Cu (Figures 5A and 5B, respectively). The average diameter of nanospheres is 56.3 ± 12.1 nm to 70Ni and 46.4 ± 5.6 nm to 70Cu. With 100 mTorr of acetylene, a nano spongy structure is obtained with some nucleation sites tubes arranged along the surface of the films, as showed in Figure 6. At the pressure of 79 mTorr of acetylene, a huge number of carbon nanotubes are formed as shown in Figure 7. They probable grow from the bottom of the nanoholes and more than one nanotube grows from each nanohole. From the Figure 7 we can readily suppose that as each nanotube grows, it laterally touches other nanotubes rom the same hole, forming nanotube bundles which in turn touch other adjacent bundles, resulting, in some regions of the film, in a very dense sponge-like structure. The average height of nanofibers is 343.0 ± 32 nm to 79Ni sample and 381.4 ± 16.5 nm to 79Cu sample. Average diameter of nanofibers estimated in the half of the tubes is 46.7 ± 13.6 nm to 79Ni and 27.9 ± 7.9 nm to 79Cu sample, respectively. It is interesting to compare our results with those of other investigators in which PECVD was also used to obtain carbon nanostructures and external power sources were used for substrate heating. Hoffman at al. [25] have grown carbon nanofibers over a Si substrate decorated with micrometric patches of 8.0 nm thick Ni films by a lithographic process. They used an external power source to heat the substrate at 120 °C. Wang et al. [11] Obtained carbon nanofibers at 140 °C over 8.0 nm Ni films, also using an external power source for heating. In
7
their work the nanofibers have length about 600.0 nm, but presented rough edges and pores. Our method of nanostructure synthesis, therefore, has the advantage that no thermal treatment of the catalysts before deposition is required, and there is no need of an external power source for substrate heating. These simplifications are of interest in the technological point of view. According to the SEM images, particularly in Figure 7, due to no pre-treatment of the film catalyst, it is possible to estimate that the nanofibers showed in our work have a “base growth” mechanism [11, 19, 23]. Wang and Moore [11] showed that due to the deposition process of Fe and Ni catalyst nanoparticle by sputtering, the base growth process is favored as the catalyst nanoparticle has greater adhesion to glass or Si substrate due to the formation of bonds between them. As the growth process takes place at temperatures of about 100oC , the nanoparticle does not melt and nanofibre growth occurs by diffusion and dissociation of carbon in the metal, this occurring at lower energy because the carbon is largely dissociated in plasma gas phase. As can be seen in SEM images that the nanofibers have a conical shape of approximately 500.0 nm in height with a base diameter smaller than top, suggesting that the break of growth occurs on the substrate (Ni and Cu nanoparticle). In this case, the mechanism of nanofibre growth is initiated by the adsorption of the monomer on the surface of the film catalyst, which is a continuous film, as can be seen in the AFM images of the catalysts films (Figure 2), which hinders the migration of the particle to the top of the nanofiber, remaining fixed on the base, in the substrate. With the nanofibers growth, the diffusion of carbon through the metal surface decreases along with the minimum energy for catalysis, prevented by the barrier created by nanofibers, this barrier increases with the height of the conical shape nanofibers generated culminating in the disruption of growth observed in the base.
8
Another factor that may contribute to the stop of nanofiber growth is the low diffusion of carbon or carbon-containing species generated in the plasma discharge on the surface of the metal catalyst due of certain quantity of amorphous carbon structure formed between nanofibers during the plasma discharge. As can be observed in Figure 7, the fibers and the amorphous carbon film grows and forms a dense structure that can act as a barrier to carbon permeation, stopping fiber growth and contributing to the smaller diameter of the fibers at their basis. The carbon nanostructures were further investigated by Raman spectroscopy. Figure 8 shows a Raman spectrum of sample 79Ni which is representative of the spectra of 79Cu sample. Pronounced peaks are seen, readily identified as the D (disorder) and G (graphitic) bands, at 1360 and 1610 cm-1, respectively. The high intensity of the G band is taken as an evidence of the large sp2 carbon density in the fibers, while the overlapping of the G and D bands is interpreted as a mixture of sp2 and sp3 bonding states. The carbon nanofibers showed in image (Figure 7A), which are basically defective nanotubes have sp2 carbons in its crystal lattice demonstrated in 1587 cm-1 (G band) with an higher intensity than the shoulder formed by the D band at 1370 cm-1 indicating a combination of sp3 carbons in the nanostructure. But in this case, the linewidths is large to compare with crystalline 2D graphite due presence of residual amorphous carbon. In fact our nanofibers exhibited a mixture between crystalline and amorphous carbon in the structure.
4. Conclusions In this work we have shown that the use of microwave-assisted PECVD can be a successful technique for the synthesis of carbon nanotubes. The various acetylene partial pressures we have used facilitate the observation of the morphology of the carbon nanoholes and nanotubes in their various stages of formation.
9
The very well-defined images of the carbon nanotubes and of their coalescence provide strong evidence for the base growth mechanism proposed in other investigations. From the Raman spectrum, the sp2-nature of the carbon nanotubes and the existence of mixed sp2-sp3 bond states was clearly evidenced. The PECVD system we have used is rather simple and inexpensive. Furthermore, the procedure adopted does not require any elaborate previous substrate treatment such as thermal annealing or holographic processes. Acknowledge The authors thank to CNPq and FAPEMIG for the financial support to this work.
References
1.
Kroto, H W., Heath J. R., O’brien, J. C., Curt, R. F. & Smalley, R. E. C60:Buckminsterfullerene. Nature 318, 162–163 (1985).
2.
Iijima, S. Helical microtubules of graphitic carbon. Nature 354, 56–58 (1991).
3.
Herbst, M. H. Tecnologia dos Nanotubos de carbono: tendências e perspectivas de uma área multidisciplinar. Química nova 27, 986–992 (2004).
4.
Stankovich, S. et al. Graphene-based composite materials. Nature 442, 282–286 (2006).
5.
Grill, A. Diamond-like carbon: state of the art. Diamond and Related Materials 8, 428– 434 (1999).
10
6.
Melechko, a. V. et al. Control of carbon nanostructure: From nanofiber toward nanotube and back. Journal of Applied Physics 102, 074314–2–074314–7 (2007).
7.
Ren, Z. F. Synthesis of Large Arrays of Well-Aligned Carbon Nanotubes on Glass. Science 282, 1105–1107 (1998).
8.
Pham-Huu, C. et al. About the octopus-like growth mechanism of carbon nanofibers over graphite supported nickel catalyst. Journal of Catalysis 240, 194–202 (2006).
9.
Chen, Y., Guo, L. ., Johnson, D. . & Prince, R. . Plasma-induced low-temperature growth of graphitic nanofibers on nickel substrates. Journal of Crystal Growth 193, 342–346 (1998).
10.
Choi, W. S., Choi, S.-H., Hong, B. & Lee, J.-H. Growth of carbon nanotubes on glass substrate by MPECVD. Materials Science and Engineering: C 26, 1215–1218 (2006).
11.
Wang, H. & Moore, J. J. Low temperature growth mechanisms of vertically aligned carbon nanofibers and nanotubes by radio frequency-plasma enhanced chemical vapor deposition. Carbon 50, 1235–1242 (2012).
12.
Hofmann, S., Ducati, C., Robertson, J. & Kleinsorge, B. Low-temperature growth of carbon nanotubes by plasma-enhanced chemical vapor deposition. Applied Physics Letters 83, 135–137 (2003).
13.
Löffler, R. et al. Optimization of plasma-enhanced chemical vapor deposition parameters for the growth of individual vertical carbon nanotubes as field emitters. Carbon 49, 4197– 4203 (2011).
11
14.
Faria, A. & Gomes, N. Filmes de nanodiamantes para aplicações em sistemas eletroquímicos e tecnologia aeroespacial. Química nova 29, 129–136 (2006).
15.
Moshkalev, S. a. & Verissimo, C. Nucleation and growth of carbon nanotubes in catalytic chemical vapor deposition. Journal of Applied Physics 102, 044303 (2007).
16.
Guo, T. et al. Self-Assembly of Tubular Fullerenes. Journal of Physics and Chemistry 99, 10694–10697 (1995).
17.
Anderson, T. T. et al. A plasma arc reactor for fullerene research. Review of Scientific Instruments 65, 3820 (1994).
18.
Lee, B. J. & Kim, J. Recent Progress in the Synthesis of Porous Carbon Materials. advanced materials 18, 2073–2094 (2006).
19.
Dresselhauss, M. S., Dresselhauss, G. & Avouris, P. Carbon Nanotubes: Synthesis, Structure, Properties and Applications. (Springer: 2001).
20.
Hoshi, F. et al. Field emission and structure of aligned carbon nanofibers deposited by ECR-CVD plasma method. Diamond and Related Materials 10, 254–259 (2001).
21.
Kolahdouz, Z. et al. Substrate engineering for Ni-assisted growth of carbon nano-tubes. Materials Science and Engineering: B 177, 1542–1546 (2012).
22.
Wang, Z., Shoji, M. & Ogata, H. Carbon nanosheets by microwave plasma enhanced chemical vapor deposition in CH4–Ar system. Applied Surface Science 257, 9082–9085 (2011).
12
23.
Baker, R. T. K., Barber, M. A., Harris, P. S., Feates, F. S. & Waite, R. J. Nucleation and Growth of Carbon Deposits from the Nickel Catalyzed Decomposition of Acetylene. Journal of catalysis 62, 51–62 (1972).
24.
Boskovic, B. O., Stolojan, V., Khan, R. U. a, Haq, S. & Silva, S. R. P. Large-area synthesis of carbon nanofibres at room temperature. Nature materials 1, 165–8 (2002).
25.
Hofmann, S., Kleinsorge, B., Ducati, C. & Robertson, J. Controlled low-temperature growth of carbon nanofibres by plasma deposition. New Journal of physics 5, 153.1–153.3 (2003).
26.
Tuinstra, F. Raman Spectrum of Graphite. The Journal of Chemical Physics 53, 11261130 (1970).
27.
Dresselhaus, M. S., Dresselhaus, G., Saito, R. & Jorio, a. Raman spectroscopy of carbon nanotubes. Physics Reports 409, 47–99 (2005).
28.
Ferrari, a. & Robertson, J. Interpretation of Raman spectra of disordered and amorphous carbon. Physical Review B 61, 14095–14107 (2000).
29.
Ayoub, S. & Beaulieu, L. Y. The surface morphology of thin Au films deposited on Si(001) substrates by sputter deposition. Thin Solid Films 534, 54–61 (2013).
13
TABLE CAPTION Tabel I: Deposition parameters and catalyst metal for each sample. For all samples: deposition time of 20 minutes; pressure of the C2H2-Ar mixture constant at 210 mTorr; maximum substrate temperature: of 120 ºC.
FIGURE CAPTIONS Figure 1: Schematic drawing of PECVD system. Figure 2: AFM images of Si waffer and Cu and Ni catalyst films used . Figure 3: SEM images of structures deposited with 34 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 4: SEM images of structures deposited with 64 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 5: SEM images of structures deposited with 70 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 6: SEM images of structures deposited with 100 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 7: SEM images of structures deposited with 79 mTorr of acetylene upon 3nm Ni (A) and Cu (B) film respectively. Figure 8: Raman spectra of nanofibers deposited with 79 mTorr of acetylene upon 3nm Ni.
14
TABLE 1 Sample
Catalyst
Catalyst
Argon
Thickness pressure
Acetylene
Structure
pressure
observed
(nm)
(mtorr)
(mtorr)
34Ni
Ni
3
110
34
Nanocones
34Cu
Cu
3
110
34
Nanosponge
64Ni
Ni
3
70
64
Nanoholes
64Cu
Cu
3
70
64
Nanoholes
70Ni
Ni
3
70
70
Nanospheres
70Cu
Cu
3
70
70
Nanospheres
79Ni
Ni
3
60
79
Nanofibers
79Cu
Cu
3
60
79
Nanofibers
84Ni
Ni
3
50
84
Nanoholes
84Cu
Cu
3
50
84
Nanoholes
15
FIGURE 1
MICROWAVE SOURCE
NEEDLE VALVES
QUARTZ CHAMBER
VACUUM PUMP
GAS SOURCE
16
FIGURE 2
17
FIGURE 3
A
B
18
FIGURE 4
A
B
19
FIGURE 5
A
B
20
FIGURE 6
A
B
21
FIGURE 7
A
B
22
FIGURE 8
G Intensity (arb. units)
D
1000
1200
1400
1600
1800 -1
Wavenumber (cm )
2000
